Intel releases 35W Intel Core i9-11900T and Core i7-11700T Rocket lake procs (with 115W PL2)

Published by

Click here to post a comment for Intel releases 35W Intel Core i9-11900T and Core i7-11700T Rocket lake procs (with 115W PL2) on our message forum
https://forums.guru3d.com/data/avatars/m/273/273678.jpg
TDP only accounts for thermal design during real world usage, the cooler system needs to displace and diffuse xxxw amount of power. the cpu itself can be using 95w but consumes from the socket and needs a cooler of 130w.
The TDP is typically not the largest amount of heat the CPU could ever generate (peak power), such as by running a power virus, but rather the maximum amount of heat that it would generate when running "real applications." This ensures the computer will be able to handle essentially all applications without exceeding ...
100% gaming loads, are not the same as 100% Folding loads.
data/avatar/default/avatar08.webp
Astyanax:

TDP only accounts for thermal design, the cooler system needs to displace and diffuse xxxw amount of power. the cpu itself can be using 95w but consumes from the socket and needs a cooler of 130w.
if it's using 95W, then it's consuming 95W as a perfect 100% efficient heater and its dissipating 95W of heat why would it NEED 130W cooler? because of PSU inefficiency? If that's what you mean - then NO 🙂
https://forums.guru3d.com/data/avatars/m/273/273678.jpg
Noisiv:

why would it NEED 130W cooler?
because its not a 100% efficient heater, this is a illfounded theory based on the idea that the processor has the same resistance values the entire way through and ignores entropy (state change)
data/avatar/default/avatar31.webp
Astyanax:

because its not a 100% efficient heater, this is a illfounded theory based on the idea that the processor has the same resistance values the entire way through.
No... it's conservation of energy - what comes in must come out. And what's at the exit end: heat, and some EM radiation. And some infitesimaly small work done on irreversible effects like silicone degradation. So OK... its not 100% efficient heater - it's only 98% efficient. 100 joules go in and only 98 joules are dissipated as a heat. Why would you need 130 joules to be dissipated?
https://forums.guru3d.com/data/avatars/m/273/273678.jpg
Noisiv:

100 joules go in and only 98 joules are dissipated as a heat. Why would you need 130 joules to be dissipated?
the cpu decided to boost and tdp goes out the window. obviously this ignores the cpu consumes x but tdp needs to be y, but the point needs to be made TDP is a worthless metric except when shopping for a AIO or heatsink.
data/avatar/default/avatar13.webp
Astyanax:

the cpu decided to boost and tdp goes out the window. obviously this ignores the cpu consumes x but tdp needs to be y, but the point needs to be made TDP is a worthless metric except when shopping for a AIO or heatsink.
Well yes. Not least of all because that's how Intel has always been treating TDP. TDP, which up to recently has more-less coincided with "power consumption" Ok I get what you mean by 130W. Nothing to do with ill-founded resistance theories, but with the fact that 95W refers only to base frequency, and not to Boost. And then there is AVX which will put it even above 130W if run without negative offset. My 8700k is around 75 watts in Prime95. At stock frequencies, undervolted. But without AVX. 110-120W with full speed AVX. 0 negative offset. So easily 150W+ if it wasn't for undervolting.
data/avatar/default/avatar37.webp
Fox2232:

So, there may be heatsink + fan capable to dissipate 130W of heat while its temperature and temperature of cooled objects remains stable. Yet one such scenario may be: Environment temperature 20°C, CPU temperature 65°C, heatsink's center of mass temperature 45°C. Or: Environment temperature 20°C, CPU temperature 85°C, heatsink's center of mass temperature 60°C. 1st scenario is cooling solution that requires small delta T to deliver 130W cooling capacity. 2nd scenario requires bigger delta T to deliver same cooling capacity. Yet both can claim ability to dissipate 130W of heat.
No shit. That's because the intricacies of the entire process that Intel takes into consideration are way above our paygrade. Terms that we are discussing like TDP are outward communicated terms, terms for general public and system builders. God know what (else) they're using in their labs and in internal documents or how anal they have to be when internally defining "simple" quantities like Power Consumption. is what I hinted to you already.
Fox2232:

And that brings rather simple definition for cooling solution requirement which nobody uses: CPU die has certain target operational temperature "Tc". It has IHS and indium in between. There is bound to be temperature difference "Tz" between CPU die and surface of IHS. Cooling solution adequate for CPU of specific "Px" power draw must be able to dissipate "Px" watts while it keeps IHS at temperature "Ti" under ambient temperature "Ta". Where "Ti" is target temperature of CPU "Tc" - "Tz".
What makes you think that's NOT how it's internally done, tested and sampled? What makes you think that's NOT how at the end of that process they came up with 95W for general public. General public who doesn't need to be bothered and confused with pressure and dew point and whatnot.
data/avatar/default/avatar31.webp
Fox2232:

Once more, Power consumption is not something you or intel can bend. They can bend a lot of things, but not SI and direct derivatives.
You are talking to a physicist. What SI are we bending. Off you go again to one of your trips... jesus.
https://forums.guru3d.com/data/avatars/m/273/273678.jpg
Noisiv:

And then there is AVX which will put it even above 130W if run without negative offset.
AVX consumption is nuts o.o, i was using avx optimized folding cores recently and the 1680v2 was pulling 180w, meanwhile cinebench r15 was close to 140.
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
Quite an interesting conversation you had there, and it has also had real world consequences. I recently watched a Youtuber (Dawid Does Tech Stuff) review some more or less terrible pre-built "gaming" PCs. He noticed the CPU coolers can abysmal. He was even in contact with one manufacturer because the marketing material of the pre-built had a consideably beefier cooler than the one he found inside his PC (it turned out the better looking cooler was only for illustration purposes, haha). The manufacturer told him there was no problem and the pitiful cooler was perfectly within specs. The Youtuber demonstrated how with a real cooler the gaming performance was better, surprise surprise. After reading you guys' discussion above, I can understand how the manufacturer had the guts to say the toy cooler was totally fine and okay. And why it did hamper performance despite being okay on paper.
https://forums.guru3d.com/data/avatars/m/227/227994.jpg
Hilbert needs to review one, see how it compares in gaming vs the 11400, 11600 and 11700.
data/avatar/default/avatar04.webp
Fox2232:

Then as one, you should be able to get from SI units to W. Or is "Power consumption" something else in your dictionary? Because you implied that intel is free to bend likes of "Power Consumption".
What I implied is that when dealing in practice with low-error-tolerance product on industrial level - a host of not trivially obvious issues needs to be taken into consideration. And so even with the most basic quantities like Power Consumption. Is why I called it "simple" - an obvious sarcasm at the fact that when setting requirements for a billion dollar product - nothing is simple. You otoh are hammering that dead horse of "Power Consumption is power consumption", and accusing Intel and me that we are saying differently LOL>
data/avatar/default/avatar40.webp
Fox2232:

Intel is bending what they can by leaving relevant information out. Or are using definitions separately from products, so they have legally covered backs and as such can be false. Yet, Mailman is weirdly defending them to death. And even puts his occupation at front line.
If what i am doing is defending Intel to death (LOL OK), then WTF are these walls-of-text that you're choking my ISP and me with? Is that not you killing us to death, you boa constructor
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
TheDeeGee:

Hilbert needs to review one, see how it compares in gaming vs the 11400, 11600 and 11700.
Like someone already said, he would need to test it with a cooler found from a cereal box and some mighty 350W PSU, with a mobo that's fitting for the theme. That's apparently the intended working environment for this particular line of CPUs.
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
Fox2232:

Isn't it bit overkill? I've built system with old Ryzen 5 2400G (65W) and tiny 150W PSU was more than enough.
He uses an RTX 2080 Ti and/or 3090 in the reviews, so I reckon a little bit more juice is needed. 350W might be a good compromise, which makes the CPU and GPU fight tooth and nail over the last watts, as intended.
data/avatar/default/avatar17.webp
interesting talk here. As usual we can see the AMD defenders, and the INTEL technical fan The fisrt one talk about IPC, WATT, TDP, 14nm again LOL...that all better yes, amd won The second about about IHS thermal transfert, frequency, fab process, and end user fact, like prebuild pc gamer matter. i hate being in a fanclub war but i love play with a cpu, underclock, overclock, put the IMC into the deep with some nice bdie, put a cooper IHS on, add 1,5KG of cooper cooler. I dont care about TDP, only cpu fun, tune for nice temp, nice perf! a lot of CPU since 7 years ago are granted you to work/game confortably. Want better perf ? go for 4 channel memory cpu...or something here we have just another cpu, with a missing old Turbo button on the case